Intelligent optimization and performance prediction of lattice-based protective structures for photovoltaic power stations

Photovoltaic (PV) power stations are susceptible to impacts from hail or debris, which threaten the power output and the structural integrity of the PV equipment, making protective measures an imperative. Though lattice structures have shown outstanding performance in energy adsorption in other areas, their performance in PV protection, and in balancing the impact resilience with optical transmittance remains uninvestigated. Moreover, there is no systematic framework that can enhance the lattice mechanical performance, the light penetration effect, and the material utilization efficiency. To address these challenges, this study proposes an integrated framework for intelligent optimization and performance prediction of lattice-based protective structures in PV power stations. Specifically, this study innovatively introduced lattice metamaterials into the protection of PV power stations. Based on the multi-scale design and relative density equations, and the impact resistance mechanism, an integrated framework that combined the particle swarm optimization (PSO) algorithm for multi-objective optimization and a hybrid neural network for performance prediction was proposed. Experiments and real-world case studies in three countries showed that the integrated framework increased the energy adsorption efficiency by 20%–30%, reduced the facility weight by 35%–40%, and maintained the optical transmittance at 83%–87%; the framework also reached a high accuracy, with the mean error remaining below 8%. Real-world case studies showed that the framework remained robust across varied climate conditions. The present work provides an efficient solution to intelligent design and performance prediction of lattice-based protective structures, and contributes to the improved impact resistance and operational reliability of PV infrastructure and other renewable energy infrastructure under varied environmental conditions.